It was great to read that so many people start using AddaxAI. As node manager for the Global Biodiversity Information Facility - gbif.org - I was wondering how many people start sharing the data with GBIF. This would be a great opportunity to increase knowledge and data about the distribution and occurrences of species. Please reach out if there is an interest in sharing data.
My understanding is that many (most)? AddaxAI users review their images in Timelapse, and the ability to upload to GBIF is one of the main reasons that Timelapse added support for Camtrap DP export a couple months ago. You can read about the details in Timelapse Metadata Guide (see page 10).
I wouldn’t say it’s zero extra work for users to export to Camtrap DP - or to push to GBIF after Camtrap DP export - but having it supported directly in Timelapse is a big step forward IMO, since it’s no longer something where users would have to write custom code, they just have to fill in a bunch of extra metadata. That’s the stuff they would have to do before public release regardless of the implementation, i.e. metadata they have already, but might not normally attach in a standardized in the course of their workflow, like location, and species names in a standard taxonomy. So for users who either just want to contribute data or whose funding requires it, it’s still not zero cost to make data public, but it’s hopefully a big step forward.
It would be great if GBIF leadership and node managers could try this out in Timelapse, provide feedback to Saul (who develops Timelapse), and evangelize.
2 Likes
@nielsraes Love the idea to make data sharing more accessible!
I agree with @agentmorris that the easiest way of sharing data with GBIF would be via Timelapse. Currently, standalone AddaxAI is not the best way for that. A future AddaxAI release will also offer CamtrapDP format exports, but that point is still far on the horizon…
Many users use Timelapse for reviewing the model predictions. I couldn’t say how much exactly, but based on user feedback, I would say that about 40% - 50% of the users use AddaxAI only to predict and then import the results to Timelapse for further processing. So Saul would be the person to reach out to.